8
BOOK REVIEW Michael Tooley, Time, Tense and Causation, Clarendon Press, Oxford, 1997, XVI + 399 pp., GBP40.00 In this work, Michael Tooley sets out and defends an original and thought- provoking account of the nature of time and its connection with causation. Tooley puts a high premium on clarity and providing arguments for his views where he can. In addition, he is careful to distinguish his view from similar rivals, and to make clear what his criticisms of the rival views are. This commitment to clear and detailed argumentation does, of course, mean that it is easy for the reader to spot the assumptions which they find implausible or do not agree with, and many will find assumptions or premises in this book they find implausible (which, I suppose, is the peril of defending an unorthodox approach). Tooley defends a view in his book which “lies somewhere between traditional tensed and tenseless approaches” (p. 381): on the one hand he accepts that tenseless facts are ontologically basic, and that semantically tensed language can be analysed in tenseless terms, while on the other hand he rejects what he calls the ‘static’ conception of time, that change is to be explained in terms of one set of facts being located at one time and another being located at a later time: Tooley wishes to defend a ‘dynamic’ concep- tion of time, where what is actual as of one time, and what propositions are true as of one time, is different from what propositions are actual as of another time, or true as of another time. Let me first briefly outline the structure of Tooley’s view. In a (perhaps over-compact) nutshell, Tooley’s view is that “the past and present are real, but the future is not” (p. 375). The world is dynamic, in the sense that “the totality of temporal facts, or states of affairs, is different at different times” (p. l4) – states of affairs are constantly coming into existence, as the world grows through time. Tooley reaches this conclusion through considerations of the nature of causation: he claims that causation and probability are connected in various ways (again to oversimplify, standardly the posterior probability of an effect is raised by its cause and the causal law together, but the posterior probability of a cause is not raised by its effect and the causal law together), and Tooley argues that those who accept a ‘static’ ac- Erkenntnis 50: 141–148, 1999.

Michael Tooley, Time, Tense and Causation

Embed Size (px)

Citation preview

Page 1: Michael Tooley, Time, Tense and Causation

BOOK REVIEW

Michael Tooley,Time, Tense and Causation, Clarendon Press, Oxford,1997, XVI + 399 pp., GBP40.00

In this work, Michael Tooley sets out and defends an original and thought-provoking account of the nature of time and its connection with causation.Tooley puts a high premium on clarity and providing arguments for hisviews where he can. In addition, he is careful to distinguish his view fromsimilar rivals, and to make clear what his criticisms of the rival views are.This commitment to clear and detailed argumentation does, of course,mean that it is easy for the reader to spot the assumptions which theyfind implausible or do not agree with, and many will find assumptions orpremises in this book they find implausible (which, I suppose, is the perilof defending an unorthodox approach).

Tooley defends a view in his book which “lies somewhere betweentraditional tensed and tenseless approaches” (p. 381): on the one hand heaccepts that tenseless facts are ontologically basic, and that semanticallytensed language can be analysed in tenseless terms, while on the other handhe rejects what he calls the ‘static’ conception of time, that change is to beexplained in terms of one set of facts being located at one time and anotherbeing located at a later time: Tooley wishes to defend a ‘dynamic’ concep-tion of time, where what is actual as of one time, and what propositionsare true as of one time, is different from what propositions are actual as ofanother time, or true as of another time.

Let me first briefly outline the structure of Tooley’s view. In a (perhapsover-compact) nutshell, Tooley’s view is that “the past and present are real,but the future is not” (p. 375). The world is dynamic, in the sense that “thetotality of temporal facts, or states of affairs, is different at different times”(p. l4) – states of affairs are constantly coming into existence, as the worldgrows through time. Tooley reaches this conclusion through considerationsof the nature of causation: he claims that causation and probability areconnected in various ways (again to oversimplify, standardly the posteriorprobability of an effect is raised by its cause and the causal law together,but the posterior probability of a cause is not raised by its effect and thecausal law together), and Tooley argues that those who accept a ‘static’ ac-

Erkenntnis50: 141–148, 1999.

Page 2: Michael Tooley, Time, Tense and Causation

142 BOOK REVIEW

count of time, according to which there is no gulf between the past and thefuture, cannot explain the probability- related asymmetry between causesand effects (p. 109). However, Tooley claims his ontological asymmetrycan explain the asymmetry. Tooley’s conclusion at this point is that forthere to be causation, the world must be a dynamic one of the sort thatTooley favours, or very similar.

To go with his metaphysical view, Tooley has some unorthodox viewson semantics. Tooley claims that there are two types of truth: truth at atime, and truthsimpliciter. Truth simpliciter is the truth which soundsmost orthodox (indeed, it seems like the sort of truth endorsed by staticapproaches): it is bivalent, it is such that statements about the future aretrue or false in the usual sort of way, and which propositions are true sim-pliciter does not seem to depend on what time it is. Tooley also endorsesthe notion of truth at a time. Truth at a time is different: standard contingentpropositions may be true at one time but not true at another (where this isnot justsentences, or propositional functions, but propositions). Roughly,the idea is that more and more claims are coming to be true: it is truetoday, but not yesterday, that I write this sentence. It is not yet true that Iwill read this sentence tomorrow (even though this will have become trueby the time you see these words). Truth at a time is not bivalent: as well astrue and false claims, contingent claims about the future get the truth value‘indeterminate’. To go with his two kinds of truths, Tooley recognises twokinds of existence (or two kinds ofbeing actual, to use his expression):there is being actualsimpliciter (and in this sense future objects and factsare actual), and there is being actual as of a time: as of now, the sentences Iwrite tomorrow do not exist, as of the day after tomorrow, they do. Tooleyresists the attempt to reduce the latter into the former: neither reduce to theother (see e.g., pp. 40–2).

Tooley then goes on to argue that the direction of time is given by causa-tion. (Indeed, he argues for a slightly stronger claim – not just the direction,but the ordering of temporal points is given by their causal ordering). Too-ley thus presents us with a picture of a world where causation and time areclosely intertwined, and an account in which the world is dynamic, withevents coming into existence and in which the future is unreal. BecauseTooley is committed to a line between those states of affairs which areactual as of now and those which are not, he is committed to an absolutepresent, and absolute simultaneity, as he admits (p. 338). This means, inturn, that he is commits himself to the falsehood of the Special and Generaltheories of Relativity. Tooley bites this bullet in his penultimate chapter,arguing that the approach he favours provides the basis of a theory superior

Page 3: Michael Tooley, Time, Tense and Causation

BOOK REVIEW 143

to Special Relativity in any case. I shall return to discussion of particularslater, but first let me remark on the general methodological approach.

Tooley’s approach to establishing metaphysical conclusions about thenature of time and causation is an odd one: he proceeds by means ofconceptual analysis of the notions of causation and temporal concepts inorder to attain conclusions like the impossibility of backwards causation,the causal nature of time, and indeed a large bulk of his view. Tooley isnot always explicit that what he is doing is conceptual analysis: he usuallyjust talks of analysing the various things (causation, the direction of time,counterfactuals, or whatever). However, there is a good deal of evidencethat this ‘analysis’ is conceptual analysis. Tooley repeatedly describes theresults of his analysis as revealing conclusions about what is logically pos-sible and what is not; his practice of employing thought experiments toestablish the falsity of rival positions by showing there are (conceptually)possible situations in which they fail, his occasional talk of the need foranalysis because of the non-basicness of certain concepts, and the factthat the truths which flow from the analyses are sometimes described asanalytic (see e.g., p. 257) suggest that his ‘analysis’ is meant to be concep-tual analysis. The rationalist project of discovering the nature of the worldthrough reflection on our concepts is not as common as it once was, butTooley has the courage of his convictions, as demonstrated by his rejectionof Special Relativity on philosophical grounds (and conceptual groundsat that, as I hope to demonstrate below). Tooley’s reliance on conceptualanalysis in these contexts will arouse a great deal of suspicion. There arethose in these Quinean times who doubt that there is any scope for makingany philosophically interesting discovery by means of conceptual analysis.As well as this general suspicion, though, there is a particular suspicion,which I must admit I myself feel, about attempting to draw metaphysicalconclusions out of conceptual hats. Even if it could be shown, after all, thatcausation can only take place in a dynamic world, on pain of conceptualincoherence, this would not settle the metaphysical questions decisively.For even if we grant Tooley’s conceptual conclusions, there is still noguarantee that there is anycausationor time in Tooley’s rich sense – whatimmediately would become a more pressing doxastic possibility, espe-cially for friends of a static approach, is that the world has only causation∗and time∗, phenomena which are very similar to the way causation or timewould be, except that they are found in a static universe. It is true that todeny that our world lacks any causation or any temporal features soundsutterly ridiculous: but if the denial is followed up with the claim that theworld contains things, causation∗ and time∗, indistinguishable from cau-sation and time by everyone except trained philosophers, and that these

Page 4: Michael Tooley, Time, Tense and Causation

144 BOOK REVIEW

causation∗ and time∗ are the things in the world we mistakenly take to becausation and time, then the denial will seem to most non-philosophers atleast as being not much more important than a semantic quibble. Thereis always a tradeoff in employing conceptual analysis in inquiry: one canemploy demanding concepts, but run a greater risk of nothing answering tothem, including the phenomena which are commonly supposed to answerto them, or one can employ less demanding concepts, running much lessrisk of their falling to have application, but at a cost to how interesting theconceptual connections between them turn out to be. If Tooley shows usthat our concepts of time and causation are so intertwined that there canbe either only if the world has a previously unsuspected structure then hehas shown us something very interesting: but he would not have therebyshown us very much about whether the world does have such a structure,rather than the possibility that it only contains things much like causationand time which do much the same jobs, and which do not require a worldof the sort Tooley believes in.

I cannot deal in depth with every issue Tooley addresses in this book(indeed, some valuable material such as his critiques of rival dynamicviews of time merit discussion but will receive no more than this mention),so I will focus in this review on what seems the most startling claim inthe book: Tooley’s putative proof that Relativistic theories, at least in theirstandard formulations, cannot be correct. What is interesting about thisproof is that the premises are argued for through conceptual analyses, asI shall try to demonstrate. If I am right, Tooley has produced a chain ofreasoning which purports to show ona priori grounds that, for example,the Special Theory of Relativity is logically false. It is one thing to think ascientific theory false: it is quite another to think that one can tell from thearmchair that Relativity theories are false.

Tooley’s main argument from the nature of causation to the conclu-sion that there must be a dynamic world in which there is an ‘ontologicalgulf’ between the past and present, on the one hand, and the future onthe other, turns on four ‘postulates’ connecting the notion of causation (or,more strictly, causal law) to ‘logical probabilities’. Logical probability isa crucial hinge in Tooley’s argument, but he is not very forthcoming, inthis book at least, about what account he wishes to offer of it. These fourpostulates are taken from a previous book of Tooley’s, his 1987Causation.In this book it is clear that these postulates are attempting to capture truthsabout the concept of causal laws (or which are true in virtue of the con-cept): they form part of an analysis of causal laws (p. 256 of Tooley 1987).In addition, the relationship of causal laws to logical probabilities wouldpresumably bea priori if logical probabilities and the rules for updating

Page 5: Michael Tooley, Time, Tense and Causation

BOOK REVIEW 145

them are givena priori, as is often assumed (Logical probabilities are amatter of logic, after all). The postulates together provide for an asym-metry in the calculation of posterior probabilities of causes and effectsgiven information about the existence of a causal laws. The postulates arequestionable (e.g., they are used by Tooley to rule out the possibility ofcausal loops, but I foundmodus tollensmore tempting thanmodus ponensin this regard, being far from convinced that causal loops are impossible),but let me for the sake of the argument grant to Tooley that they are correct,and provide part of the conceptual analysis of the concept of a causal law.The next step remains objectionable. Tooley claims that “In a static world,then, it is impossible to explain how the crucial asymmetry expressed bypostulates C1 and C4 can obtain” (pp. 109–10). However, if one assumesthe world is dynamic in the manner which Tooley supports, he claimsthat an explanation is possible (pp. 110–11). However, it can hardly bea fault of a theory that it does not offer a metaphysical explanation fora conceptual truth, which is the charge Tooley appears to be laying atthe door of the static view. Indeed, it seems to me a fault of a view tooffer a metaphysical explanation of a conceptual truth, as Tooley’s does: itshows the theory is as misguided as a theory which offered a sociologicalaccount to explain why such a high percentage of bachelors happen to beunmarried. Alternatively, Tooley may not be requiring that the static viewexplain the conceptual truth that causation has such-and-such-connectionswith probability: instead, he may be requiring that the static view providean explanation of how there could be a phenomenon with such-and-suchconnections with probability (i.e., how their metaphysics can have a placefor causation). Even here, however, it is not clear that the static approachhas no answer: after all, the principles governing logical probability willthemselves require elucidation, and the static account might account forthe conceptual truths as true in virtue primarily of the concept of logicalprobability: it is just such that it is asymmetric with respect to probabilityassignments in some situations involving information about causal laws.It is not clear exactly what Tooley is demanding of the static view here,and not clear that it cannot live up to the challenge (whatever it is): as thisargument is one of the vital linchpins in Tooley’s stand against the staticaccount (and, as it turns out, one of the vital steps in his argument againstSR), this is unfortunate. Tooley is not explicit that his conclusion is meantto have the status of an analytic truth: but given that it is offered as an ‘ex-planation’ of conceptual truths, and that it establishes, according to Tooley,that “causation cannot exist in a static world” (p. 112), the impression ofthis reader at least was that this ‘analysis’ was indeed a conceptual one.

Page 6: Michael Tooley, Time, Tense and Causation

146 BOOK REVIEW

Tooley’s conclusion, as we have seen, is that causation requires the ex-istence of a dynamic world, on pain of violation of conceptual truths aboutcausation. (Tooley does not require quite that it be his preferred dynamicmodel – but it needs to be some non-static one). The next plank in thestrand of his argument which culminates in the rejection of SR (and itscousins) is to show that the direction of time is given by the direction ofcausation (andinter alia that the ordering of time is given by the orderingof causation). This too will be a conceptual truth, by Tooley’s lights, andso combined with the first conclusion will yield the doctrine that for thereto be time at all (or at least time which is ordered), the world must bedynamic, in more-or-less Tooley’s preferred way – and it will yield thisdoctrine as a conceptual truth. In order to show that the direction of timeis the direction of causation, Tooley relies on several arguments, centredaround the need to explain certain putative conceptual truths about tempo-ral relations, putative analytic truths connecting time and causation, and anargument based on the inadequacies of other explanations of the directionof time (see Tooley: 256–58).

The details of this analysis need not concern us here, but the vital thingto notice is that it is a conceptual truth that the direction of time is analysedin terms of the direction of patterns of causal relationships. No causation,no direction of time: and indeed no ordering of time, either. Thus any worldwithout causation is a world where time has no ordering. It is not clearwhether there could be time with no ordering, but in any case this doesnot matter: for it seems clear that Special Relativity (hereafter ‘SR’) isa theory which postulates temporal relationships and temporal orderings(temporal orderings relative to a frame of reference, of course). For Tooley,causation requires, as a conceptual matter, a dynamic world where there isa clear ontological dividing line between the past and the present, on theone hand, and the future, on the other (see above): so it is a conceptualmatter, indirectly, that it is incoherent both to assert that there are temporalorderings and deny that the world is suitably dynamic, (i.e. to deny that theworld has an absolute distinction between past and future). SR asserts thefirst and denies the second, and so, by Tooley’s argument, is necessarilyfalse on conceptual grounds.

Tooley does attempt to sugar this pill by offering an alternative theorywhich he claims is more attractive than SR on scientific grounds. Too-ley’s approach, though he does not say as much explicitly, is to return toa somewhat Lorentzian theory: instead of the ether, Tooley has absolutespace (two events are simultaneous iff they are non-derivatively spatiallyrelated). The absolute reference frame is defined as being at rest rela-tive to absolute space (and is incidentally for Tooley also the reference

Page 7: Michael Tooley, Time, Tense and Causation

BOOK REVIEW 147

frame where the speed of light is constant in all directions), and it isonly in this frame of reference that measurements of velocity, rest mass,time elapsed etc. are correct. However, like the traditional Lorentzian ap-proach, because of the postulated distortion of measurement instrumentsit is observationally impossible, on the standard theory at least, to deten-nine which is that absolute reference frame. Unsurprisingly, SR and Too-ley’s neo-Lorentzianism agree on the usual experimental predictions (Za-har’s demonstration of the ‘observational equivalence’ of Lorentz’s T3 andSR demonstrates why nicely: see Zahar, 1973. “Why did Einstein’s Pro-gramme supersede Lorentz’s?”, BJPS 24: 95–123). So, given such agree-ment in predictions (in the standard areas at least), Tooley’s theory neednot be dismissed immediately on scientific grounds. Tooley does claimscientific advantages for his theory, but none look decisive – indeed inmany cases it was not clear that the putative advantages were even genuine.

Space precludes detailed discussion of the scientific merits and demer-its of Tooley’s neo-Lorentzianism, but on this topic one aspect of Tooley’sdiscussion was disappointing: an account of the success of the predictionsof General Relativity was conspicuous by its absence. If we are to followTooley in the great leap backwards to Lorentz, he needs to explain howhe hopes to succeed where the Lorentzian program foundered. A theoryexplaining the phenomena of SR by appeal to absolute ether (or absolutespace, for that matter-and given that Tooley on p. 345 calls this absolutespace the “medium of transmission” of light, the difference is even lessthan one might expect) cannot be generalised simply and attractively in amanner analogous to the generalisation from SR to the General Theoryof Relativity. But the General Theory of Relativity has a great deal oftheoretical success and a deal of experimental support – and while, as wehave seen, Tooley is not adverse to ruling out successful scientific theorieson a priori grounds, he will hopefully feel obliged to come up with analternative explanation for the phenomena predicted and explained by the(alsoa priori impossible for Tooley) General Theory of Relativity. Thisbook does not mention this, let alone offer any suggestions as to how thisproject could be carried out – but it ought to be remembered that relativistictheories of space-time have more support than that provided for SR alone.

I am sceptical that Tooley will be the spearhead of a Lorentzian revival(though were someone to attempt a non-relativistic alternative to GeneralRelativity, that would be interesting), as I am sceptical that as much meta-physics can be done through conceptual analysis as Tooley seems to think.But his book is valuable in providing a distinctive package of views ontraditional issues in the philosophy of time, and hopefully will provoke

Page 8: Michael Tooley, Time, Tense and Causation

148 BOOK REVIEW

thought both from those unsympathetic to the intuitions and argumentsrelied on as well as those who may find the approach congenial.

Philosophy Department DANIEL NOLANMacquarie UniversityNorth Ryde NSW 2109Australia